MLFLow: Uses Cases and Implementation

MLFlow: Use Cases and Implementation

ABSTRACT – Please submit form for immediate access to white paper.

With model training occurring across different environments and being done by different people, it can be hard to keep track of what model types and parameters have been done by your organization. Further, post-hoc evaluation of models can be very challenging because images for items like training curves and different scoring metrics may not be saved in a way that is easy to recall after the fact. Both of these problems, as well as model hosting and model training curves can be solved simply with an open source tool called MLFlow by Databricks.

MLFlow is an open source tool that ML and Data Science professionals need to know about as it can make many parts of their lives significantly easier with just a few lines of code. People who manage these professionals also can get a lot of value from the tool because they can easily get access to information about how the models are performing and can present these results without having to take their team away from their work.